On the Keil evaluation version there is no float supported.

But I think, the free SDCC include float support.

Even if you program in assembler, you can still link it with C libraries together.
You should write a small C progam, which call e.g. the sin() function and then look on the assembler output, how the parameter passing was done.

But the easiest way should be, write the whole program in C.


Peter